How many people died in institutions from COVID-19?

The short answer is that we don't know

During 2020, the question: how many people have died from COVID-19 in an institutional setting? dominated news coverage of deaths. Numbers as high as 81% circulated in the press, suggesting that institutions were overseeing mass deaths, the likes of which was not experienced in any other OECD country.

Those early, high estimates were probable incorrect for many reasons. The biggest one was that institutions have strict legal reporting requirements and access to testing. While we had a good idea of who was dying within institutions, we didn’t really know who was dying in the general population. As reporting protocols got better, we did get a more fulsome image of out-of-institution death. I say more with a lot of emphasis — it still isn’t enough.

Since 2020, institutions have done a lot to improve their internal infection controls, though many systemic issues persist. As a result, of deaths that have been officially linked to institutions have dropped as a percentage of overall deaths. But no one is paying attention to this any more. And so, as I still am tracking institutional deaths, I posed this question to myself on Twitter on November 23. Here’s what I found by using a bit of math and a few assumptions, with numbers updated to this week.

There have been 48,349 COVID-19 deaths in Canada. We do not have overall death totals in residential care -- LTC, retirement homes, shelters, prisons, hospitals, group homes, etc. In my own research, I have linked 21,903 deaths to 2499 facilities across Canada.

At 21,903 deaths in 2499 facilities across Canada, that is 45% of all deaths occuring in residential care. But the only provinces that report who is dying in care any longer are Quebec and, to a lesser extent British Columbia. Some provinces have given total numbers of deaths, like PEI, but that still doesn’t give us a picture of what kinds of facilities they’re in. And in Ontario, any time deaths are fewer than 5, we aren’t told what the exact number is, based on some fake concern for privacy.

Here are some theories, based on Nov. 23 data. If we take Quebec, where the reporting is best (though not so great in hospital outbreaks), 10,876 of 17,290 deaths occured in residential care. That is 63%.

While Quebec did have far more deaths, especially in care and during the first wave, we can assume that their ratio is not that far off of other locations. To test this, let's look at the City of Ottawa.

Ottawa Public Health had been, until recently (like in the past 6 weeks?), posting all deaths in all facilities from COVID. Now, they've posted aggregate numbers. While this is insufficient, it does give us an idea of what kind of deaths there have been there.

Even though I had been tracking their numbers closely, I missed a lot of deaths (for many reasons, incl back-adding deaths or deaths that didn't land on the official list for whatever reason): LTC deaths - under by 113 retirement residence deaths - under by 51 hospital deaths - under by 45

Now, the City of Ottawa has had 963 COVID-19 deaths total. 629 of these deaths occured in residential facilities. That is ... 65%. Two percent higher than Quebec's data (which might be the margin of error because of bad hospital data anyway).

That would mean that Canada likely has had 30,538 deaths in residential care ... 9000 *more* than I've been able to track, and 9000 more that provincial and territorial governments have publicly reported. Where did they happen? Who was responsible for infection control?

Now, this is an approximation. Maybe Regina is wildy different than Gatineau, or Ottawa is wildly different than Halifax. But it's as approximate as we can get. And that means that at least 30,000 Canadians died from COVID that someone had the responsibility to stop.